Objective To systematically evaluate the efficacy and safety of duloxetine versus paroxetine for adults’ depression. Methods A search was conducted in The Cochrane Library (Issue 6, 2011), Pubmed (1998 to June 2011), CNKI (1998 to June 2011), VIP (1998 to June 2011), CBM (1998 to June 2011), Wanfang database (1998 to June 2011), MEDLINE (1996 to June 2011) and Science Direct (1998 to June 2011). The randomized controlled trials (RCTs) on duloxetine versus paroxetine for adults’ depression were collected. The quality of the included trials was assessed according to the Cochrane Handbook 5.0, and the systematic analysis was conducted by using RevMan 5.0 software. Results Six RCTs involving 1 106 patients were included. The results of meta-analysis showed that: a) After eight-week treatment, there were no significant differences in the effective rate (RR=0.96, 95%CI 0.89 to 1.05, P=0.39) and the final cure rate (RR=0.99, 95%CI 0.86 to 1.15, P=0.93) between the duloxetine and paroxetine groups; b) Adverse reaction: The incidence rate of somnolence in the duloxetine group was lower than that of the paroxetine group (RR=0.63, 95%CI 0.41 to 0.96, P=0.03), oppositely, the incidence rate of abnormal ECG was higher in the duloxetine group (RR=1.91, 95%CI, 1.02 to 3.58, P=0.04). And the other common adverse reactions were not significantly different between the two groups (Pgt;0.05). Conclusion After eight-week treatment, there are no significant differences in the effective rate and the final cure rate between duloxetine and paroxetine. Duloxetine tends easily to induce the abnormal ECG compared with paroxetine.